While links between religiosity and adolescent sexual behavior have been documented, they are not necessarily well understood. This R03 is a longitudinal study using secondary data from three waves of the National Longitudinal Study of Adolescent Health (Add Health) to examine if and how religiosity influences early sexual behavior. The main objective of the study is to investigate a construct, adolescents'shared religiosity with family, comprised of adolescents'and parents'response to three religiosity items in the Add Health data: frequency of prayer, importance of religion, and beliefs in biblical inerrancy, an indicator of fundamentalism. Specifically, we propose that adolescents'shared religiosity with family will contribute to the social capital they derive from family, which will in turn contribute to less sexually risky motivation and behavior. Although our primary interest is in adolescents'early sexual motivations and behaviors, we will take advantage of Add Health's longitudinal data and explore how early experiences of shared religiosity and family social capital might influence sexual histories into late adolescence/young adulthood. Specific aims are to: (1) identify profiles of adolescents'shared religiosity with family (i.e., patterns of convergence or divergence of views and practices between adolescent and parent.), and (2) examine associations among shared religiosity with family, family social capital, motivations related to sexual activity, and actual sexual behaviors. Specifically we will examine: (a) direct effects of shared religiosity with family on sexual motivations and behaviors;(b) indirect effects through family social capital, and (c) moderators of these links such as: religious identity/denomination, participation in a faith community, family resources, strength of peer relations, and psychological resources such as efficacy and self-esteem. A third aim is to: (3) explore links between earlier (Wave 1) experiences of shared religiosity with family and family social capital, and subsequent sexual behavior one year later (Wave 2) and in late adolescence/young adulthood (Wave 3). Expected sample size is approximately 16,500 adolescent/parent pairs. Methods will include Generalized Latent Variable Modeling.